Research Article

Selecting the best probability distribution for at‑site flood frequency

analysis; a study of Torne River
Mahmood Ul Hassan1   · Omar Hayat2 · Zahra Noreen3

Received: 26 May 2019 / Accepted: 29 October 2019 / Published online: 18 November 2019
© The Author(s) 2019  OPEN

Abstract
At-site flood frequency analysis is a direct method of estimation of flood frequency at a particular site. The appropri-
ate selection of probability distribution and a parameter estimation method are important for at-site flood frequency
analysis. Generalized extreme value, three-parameter log-normal, generalized logistic, Pearson type-III and Gumbel
distributions have been considered to describe the annual maximum steam flow at five gauging sites of Torne River in
Sweden. To estimate the parameters of distributions, maximum likelihood estimation and L-moments methods are used.
The performance of these distributions is assessed based on goodness-of-fit tests and accuracy measures. At most sites,
the best-fitted distributions are with LM estimation method. Finally, the most suitable distribution at each site is used to
predict the maximum flood magnitude for different return periods.

Keywords  Flood frequency analysis · L-moments · Maximum likelihood estimation

1 Introduction insurance studies, planning of flood management and

Floods are natural hazards and cause extreme damages
throughout the world. The main reasons of floods are
extreme rainfall, ice and snow melting, dam breakage
and the lack of capacity of the river watercourse to con-
vey the excess water. Floods are natural phenomena which
cause disasters like destruction of infrastructure, damages
in environmental and agricultural lands, mortality and
economic losses. Many frequency distribution models
have been developed for determination of hydraulic fre-
quency, but none of the distribution models is accepted
as a universal distribution to describe the flood frequency
at any gauging site. The selection of a suitable distribution
usually depends on the characteristics of available data
at a particular site. We need to estimate the flood mag-
nitude at a particular site for various purposes including
construction of hydraulic structures (barrages, canals,
bridges, dams, embankments, reservoirs and spillways),
rescue operations. We come across a number of methods
which are available in the literature for flood magnitude
estimation, but at-site flood frequency analysis remains
the most direct method of estimation of flood frequency
at a particular site.
To describe the flood frequency at a particular site,
the choice of an appropriate probability distribution and
parameter estimation method are of immense importance.
The probability distributions used in this study include the
generalized extreme value (GEV) distribution, Pearson
type-III (P3) distribution, generalized logistic (GLO) distri-
bution, Gumbel (GUM) distribution and three-parameter
log-normal (LN3) distribution. These distributions are rec-
ommended for at-sight flood frequency analysis in various
countries [2, 4, 24]. Furthermore, these distributions are
most commonly traced in the hydrological literature for
at-site and regional flood frequency analysis.

*  Mahmood Ul Hassan, scenic555@gmail.com | 1Department of Statistics, Stockholm University, Stockholm, Sweden. 2Department
for Education, London, UK. 3Division of Science and Technology, University of Education, Lahore, Pakistan.

SN Applied Sciences (2019) 1:1629 | https://doi.org/10.1007/s42452-019-1584-z

Vol.:(0123456789)
Research Article SN Applied Sciences (2019) 1:1629 | https://doi.org/10.1007/s42452-019-1584-z
Cicioni et al. [3] conducted at-site flood frequency anal-
ysis in Italy by using 107 stations. They identified LN3 and
GEV as best-fitting distributions based on the Kolmogo-
rov–Smirnov (KS), Anderson–Darling (AD) and Cramer–von
Mises (CVM) goodness-of-fit tests. Saf [23] found the GLO
as the most suitable distribution for Upper-West Medi-
terranean subregion in Turkey. Mkhandi et al. [17] used
annual maximum flood data of 407 stations from 11 coun-
tries of Southern Africa to conduct the regional frequency
analysis. They identified Pearson type-III (P3) distribution
with probability weighted moment (PWM) method and
log-Pearson type-III (LP3) with a method of moment
(MOM) as the most suitable distributions for the regions.
Młyński et al. [18] identified log-normal distribution as the
most suitable for the upper Vistula Basin region (Poland).
There have been many studies in the past literature on
the comparison of various probability distributions with
different parameter estimation methods for at-site flood
frequency analysis, e.g. see [1, 5, 7, 22] (Fig. 1).
The most commonly used methods for estimation
of parameters in flood frequency analysis are the maxi-
mum likelihood estimation (MLE) method, the method of
moments (MOM), the L-moments (LM) method and the
probability weighted moments method (PWM). The MLE
method is an efficient and most widely used method for
estimation of parameters. Recently, the LM method has
gained more attention in the hydrological literature for
estimation of parameters of probability distributions. In
this research study, we used LM and MLE methods for
estimation of parameters of the candidate probability
distribution.
The methods usually use for selection of the best
distribution are goodness-of-fit (GOF) tests (e.g. Ander-
son–Darling and Cramér–von Mises), accuracy measures
(e.g. root mean square error and root mean squared per-
centage error), goodness-of-fit indices (e.g. AIC and BIC)
and graphical methods (e.g. Q–Q plot and L-moment ratio
diagram). In the hydrological literature, researchers have
used different methods in order to find the best probabil-
ity distribution. To identify the best-suited distribution
at each site of Torne River, we have used goodness-of-fit
(GOF) test and accuracy measures. The GOF tests are used
to test that the data come from a specific distribution. The
accuracy measures provide a term by term comparison of
the deviation between the hypothetical distribution and
the empirical distribution of the data. The details about
accuracy measures and goodness-of-fit test used in this
study are described in Sect. 3.
The estimation of flood frequency of the high return
period is of great interest in flood frequency analysis. The
flood frequency estimation of return periods is always
associated with uncertainties. Uncertainty in flood fre-
quency analysis arises from many sources. Uncertainties
Vol:.(1234567890)
included in water resources management can be distin-
guished in data uncertainties, structural uncertainties
and model/parameters uncertainties, see e.g. [13, 14].
Furthermore, there is uncertainty in the estimation of
flood frequency of return periods much larger than the
actual records, particularly in the type of probability den-
sity function (PDF) and its parameters. This is particularly
true on the right tail of the PDF, the region of interest for
flooding. In addition, there is uncertainty in the measure-
ments. For example, see [15] for an in-depth discussion on
epistemic uncertainty (reducible uncertainty) and natural
uncertainty (irreducible uncertainty). The flood estimation
on high return periods are always associated with high
uncertainties. In this study, we quantify the uncertainty of
a given quantile estimate for a specific fitted distribution
by using the parametric bootstrap method.
In this research paper, the flood frequency calculation,
using statistical distribution, is addressed for gauged
catchments, for which we dispose a respectively long-
term hydrological time series. The choice of an appropriate
probability distribution and associated parameter estima-
tion method is vital for at-site flood frequency analysis. The
core objective of this study is to find the best-fit distribu-
tion among the candidate probability distributions with
a particular method of estimation (MLE or LM) for annual
maximum peak flow data at each site of the Torne River by
using goodness-of-fit (GOF) tests and accuracy measures.
We are also interested to look that, is there any best overall
distribution and fitting method for these five sites of Torne
River?. Finally, to estimate the quantiles of flood magni-
tude for the return period of 5, 10, 25, 50, 100, 200 and 500
years with non-exceedance probability at each site of the
river using the best-fit probability distribution. To address
the uncertainty of flood estimations, we estimate standard
error of estimated quantiles and construct 95% confidence
interval of flood quantile for the return period using the
parametric bootstrap method. This is a first study for at-site
flood frequency analysis of Torne River.
This paper is organized as follows: Sect. 2 describes the
study area and available data for the analysis. Section 3
deals with the model description, parameter estima-
tion method and model comparison methods. Section 4
provides the results and discussion of the application of
L-moments and maximum likelihood estimation method
based flood frequency analysis of five gauging sites on the
Torne River. Finally, Sect. 5 concludes the article.
2 Study area and data
The Torne River works as a border between northern Swe-
den and Finland, with total catchment area 40157 km2 of
which 60% is within Swedish border and the remaining
SN Applied Sciences (2019) 1:1629 | https://doi.org/10.1007/s42452-019-1584-z Research Article

Table 1  Summary of Torne Station name Station no. Latitude Longitude Catchment Period for time series
River gauging sites area (km2)

Kukkolankoski Övre 16722 65.98 24.06 33,929.60 1911–2018

Pajala Pumphus 2012 67.21 23.40 11,038.10 1969–2018
Abisko 2357 68.19 19.99 3345.50 1984–2018
Junosuando 04 67.43 22.55 4348.00 1967–2018
Övre Abiskojokk 957 68.36 18.78 566.30 1985–2018

[Swedish meteorological and hydrological institute

(SMHI)]. The data of annual maximum flow of five gauging
sites of Torne River (Swedish: Torneälven) are considered in
this study. The data have been collected from SMHI (www.
smhi.se). The length of the data series varies from 34 to 108
years. The summary of Torne River gauging sites character-
istics is presented in Table 1.

3 Methodology

3.1 Candidate probability distributions

To describe the flood frequency at a particular site, the

Fig. 1  Locations of gauging stations used in this study
area is in Finland. The Muonio River, which is the biggest
contributor of the Torne River, joins shortly after Pajala
Pumphus. Another contributor river Lainio (259.74 km
long) joins the Torne river shortly after Junosuando. In
springtime, water flow is above average level, which con-
verts into flood and this flood causes the damages to the
waterfront constructions and buildings [6]. Therefore,
Torne River is frequently affected by flooding problem
selection of an appropriate probability distribution
is always important. We have considered generalized
extreme value (GEV), Pearson type-III (P3) distribution,
generalized logistic (GLO) distribution, Gumbel (GUM)
distribution and three-parameter log-normal (LN3) distri-
bution for the analysis of flood frequency at five gauging
sites of the Torne River. The probability density function
(pdf ) and quantile function y(F) of these distributions are
summarized in Table 2. These distributions are common
in the literature and are recommended distributions for
flood frequency analysis in many countries (see e.g. [2,
4, 21, 22]). We explain the detail of parameter estimation
method (MLE and LM) in the following subsection.

Table 2  Probability density and quantiles functions of the probability distributions

Distributions Probability density function f (y) Quantile function y(F)

GEV )] 1 −1 )] 1 } 𝛼[
− (− log F)𝜅
{ [ ]
1
[ (
y−𝜇 k
(
y−𝜇 k 𝜇+ 1
𝛼
1−𝜅 𝛼 exp − 1 − 𝜅 𝛼 𝜅

P3 Explicit analytical form is not available

{ }
1 (y−𝜇)
𝛽𝛼 𝛤 𝛼
(y − 𝜇)𝛼−1 exp − 𝛽

GLO )}1∕𝜅 ]−2 𝛼[

− {(1 − F)∕F}𝜅
]
[ ( )( 1 −1) ][ { ( 𝜇+ 1
1 y−𝜇 k y−𝜇 𝜅
𝛼
1−𝜅 𝛼
1+ 1−𝜅 𝛼

LN3 Explicit analytical form is not available

� � � � � ��2 �
1 1−k(y−𝜇) 1 1 1−k(y−𝜇)
√ exp − log 𝛼
− 2
− k
log 𝛼
𝛼 2𝜋

GUM
[ ( )]
1 y−𝜇 y−𝜇
exp − 𝛼 − exp − 𝛼 𝜇 − 𝛼 log (− log F)
𝛼

where 𝜇 , 𝛼 and 𝜅 are the location, scale and shape parameters of the distribution

Vol.:(0123456789)
Research Article SN Applied Sciences (2019) 1:1629 | https://doi.org/10.1007/s42452-019-1584-z

3.2 Maximum likelihood estimation (MLE) method The unbiased sample estimators of 𝛽i of the first four PWMs
for any distribution can be computed as follows
The MLE method estimates the parameters by maximizing n
the log-likelihood function of a probability distribution. b0 = n−1
∑
yj∶n
Suppose we have n independent and identically distrib- j=1
n
uted observations y1 , y2 , … , yn . Each yi has a pdf given by b1 = n−1
∑ (j−1)
y
(n−1) j∶n
f (yi ;𝝁) . Here, 𝝁 = (𝜇1 , 𝜇2 , … , 𝜇k ) is a vector of unknown j=2
n
parameters to be estimated. Then, the � log-likelihood b2 = n−1
∑ (j−1)(j−2)
y
(n−1)(n−2) j∶n
function is defined as l(𝝁) = ni=1 log f yi ;𝝁  . The maxi-
∑ �
j=3
n
mum likelihood estimate of 𝝁 is the value of the param- b3 = n−1
∑ (j−1)(j−2)(j−3)
y
eter vector 𝝁 that maximize the l(𝝁) for given data Y. We (n−1)(n−2)(n−3) j∶n
j=4
use numerical optimization methods in order to search 𝝁
which give the maximum value of l(𝝁) . Many numerical where the data ( y1∶n ) are an ordered sample in ascend-
optimization methods, e.g. Newton–Raphson method, ing order from 1 to n. The parameters with L-moments
Nelder and Mead, differential evolution, etc. are found in estimation method are obtained by equating the sample
the literature. We have used Nelder and Mead method for L-moments with distribution L-moments.
numerical optimization proposed by Nelder and Mead
[19].
3.4 Standard error of estimated parameters
3.3 Theory of L‑moments (LM)
The standard errors (SE) of estimated parameters indicate
L-moments are introduced by Hosking [9, 10], which are a measure of reliability of estimates and performance of
linear functions of probability weighted moments (PWM’s). estimation technique. In this study, we have obtained SE of
L-moments are alternative to the conventional moments, estimated parameters by Monte Carlo simulation method.
but computed from linear combinations of order statis- The description of this method is given as
tics. L-moments can be defined for any random variable Y
whose mean exists [10]. The rth-order PWM ( 𝛽r ) is defined • We use estimated parameters with MLE and LM method
as at each gauging site and draw 1000 samples of size
equal to the length of data from each probability dis-
1
tribution.
∫
𝛽r = y(F)F(y)r dFr = 0, 1, 2, … • For each simulated sample, we obtain the MLE and LM
0 estimates for the parameters of the distribution.
• For each gauging site, the standard errors are obtained
where F(y) is a cumulative probability distribution and by taking the standard deviation of these 1000 MLE and
y(F) is a quantile function of distribution. The first four LM estimates of the parameters of each distribution.
L-moments in terms of linear combination of PWM are
defined as
𝜆1 = 𝛽0
𝜆2 = 2𝛽1 − 𝛽0
𝜆3 = 6𝛽2 − 6𝛽1 + 𝛽0 3.5 Goodness‑of‑fit (GOF) tests
𝜆4 = 20𝛽3 − 30𝛽2 + 12𝛽1 − 𝛽0
The goodness-of-fit tests are used to test that the observed
The first L-moment ( 𝜆1 ) is a measure of location (mean), data follow a particular distribution. We consider the
while the second L-moment represents the dispersion. Anderson–Darling (AD) test for the study. This test is often
Finally, the L-moment ratios defined by Hosking [10] are used in flood frequency analysis and has shown good per-
given below formance in case of small sample size and heavy-tailed dis-
𝜆2
tributions [12, 20]. The test statistic for AD test is defined as
L-Coefficient of variation (𝜏2 ) = 𝜆1
𝜆3 AD = −n − S
L-Skewness (𝜏3 ) = 𝜆2 ∑n � ��
where i=1 2i−1
�
𝜆4 log(1 − F(y )) + log(F(y ))
L-Kurtosis (𝜏4 ) = 𝜆2 n n−i+1 i
where F(yi ) represents the cumulative distribution func-
tion (CDF) of the specified distribution.

Vol:.(1234567890)
SN Applied Sciences (2019) 1:1629 | https://doi.org/10.1007/s42452-019-1584-z Research Article

3.6 Accuracy measure method estimation of flood frequency of return periods. The con-
fidence interval of flood quantile gives an estimated range
In accuracy measure (AM) methods, we have used the of values which is likely to include the flood frequency of
mean absolute error (MAE), mean absolute percentage return periods. We use a parametric bootstrap method for
error (MAPE), root mean square error (RMSE), root mean estimation of standard error of estimated quantiles and
squared percentage error (RMSPE) and correlation coeffi- confidence interval of flood quantiles of return periods.
cient ( R2 ) to evaluate how adequately a given distribution This method is more precise than an asymptotic computa-
fits the observed data. These measures are defined as tion when n is small [16, p. 133]. The detail of procedure
n
involves in parametric bootstrap method is given in [16,
MAE =
1�
�F(yi ) − F(̂yi )� p. 133].
n i=1
n
100 � �� (F(yi ) − F(̂yi )) ��
MAPE =
n i=1 �� F(yi ) �
� 4 Result and discussion
�
∑n
(F(yi ) − F(̂yi ))2
RMSE = i=1 We summarized the basic statistics of five gauging sites in
n Table 5. All data on gauging sites in the table are in cubic
�
� n �
� 1 � F(y ) − F(̂y ) �2 metre per second. It is observed that all data at these sites
i i
RMSPE =� ×100 are skewed. This is a enough evidence to model the data
n i=1 F(yi )
with non-normal distribution. In flood frequency analysis,
∑n ̄ i ))2
(F(̂yi ) − F(y the basic statistical assumptions are independence, ran-
2 i=1
R =∑ domness and stationarity of the data series (see e.g. [8,
n ̄ i ))2 + ∑n (F(yi ) − F(̂yi ))2
yi ) − F(y
i=1 (F(̂ i=1 11]). The independence and randomness of the data series
at given site are tested by using correlation coefficient (r)
∑n
F(̂y )
̄ i ) = i=1 i and n represents the size of the data
where F(y n at lag-1 and Wald–Wolfowitz (WW) test, respectively. To
series. In all above accuracy measures, F(yi ) is the empiri-
check the stationarity of the data series, Mann–Kendall
cal cumulative distribution function (CDF) of the data
(MK) test has been applied. The assumptions verification
(observed ordered values) and F(̂yi ) indicates the theoreti-
results are summarized in Table 3. The results in Table 3
cal CDF of the distribution (ordered estimated values from
indicate that the data series at each gauging site of Torne
the distribution).
River are suitable for flood frequency analysis and prob-
ability density estimation.
The estimated parameters for each distribution at each
3.7 Quantile estimation
gauging site by using MLE and LM method of estimation
along with standard error (SE) are reported in Table 4. To
After selection of the best probability distribution, the
identify the best distribution at each site, we use GOF tests
main goal of flood frequency analysis is to estimate the
and accuracy measures. Each distribution with parameter
quantile yT for a return period (T) of scientific relevance.
estimation method is ranked in each GOF test and accu-
P(Y ⩾ yT ) = T1 indicates the probability of exceedance from
racy measure in Table 6. The distribution is assigned a rank
flood level yT once in T years. The cumulative probability
score between 1 and 10 in GOF test and accuracy meas-
of non-exceedance is defined as
ures, rank score 10 for the best-fitted and 1 for the worse
1 fitted distribution. The rank score scheme is based on the
F = F(yT ) = P(Y ⩽ yT ) = 1 − P(Y ⩾ yT ) = 1 −
T relative magnitude of accuracy measures and AD test P
value. The distribution with the lowest RMSE, lowest MAE,
The distribution function F(yT ) can be expressed in inverse
lowest RMSEP, lowest MAEP or the highest R2 has the high-
form as yT = y(F) , and we can directly evaluate estimated
est rank score value 10. In AD test, the distribution with
quantile yT by replacing F. Sometimes, inverse of F(yT ) does
the highest P value has the highest rank score value 10.
not exist analytically, and then, the numerical method is
The best distribution with estimation method at each site
used to evaluate yT for the given value of F. The expres-
is identified based on the total rank score in GOF tests and
sions of quantile function of candidate distributions are
accuracy measures methods. The total rank score in Table 6
summarized in Table 2. The quantile estimate for T years
indicates that GLO with MLE estimation method is best for
is calculated by substituting the value of F = (1 − T1 ) in
Junosuando. For site Pajala Pumphus and Abisko, the LN3
the expressions of quantile in Table 2. The standard error
distribution is performed better than other distributions
of estimated quantiles represents the uncertainty in the
with the LM method. The GEV and PE3 distribution with

Vol.:(0123456789)
Research Article SN Applied Sciences (2019) 1:1629 | https://doi.org/10.1007/s42452-019-1584-z

Table 3  Assumptions results Station name n r P value Mann–Kendall test Wald–Wol-

of five gauging sites of Torne fowitz test
River
Test statistics P value P value

Kukkolankoski Övre 108 −  0.08 0.23 1.85 0.06 0.86

Pajala Pumphus 50 −  0.11 0.25 0.32 0.75 0.94
Abisko 35 0.03 0.80 0.21 0.83 0.17
Junosuando 52 −  0.09 0.34 1.41 0.16 0.95
Övre Abiskojokk 34 −  0.06 0.66 −  0.62 0.53 0.33

where n represents the sample size of time series and r indicates the Pearson correlation coefficient

Table 4  Estimated parameters with MLE and LM methods

Station Probability distributions
GEV P3 GUM GLO LN3
MLE MLE MLE MLE MLE
𝜇̂ 𝛼̂ 𝜅̂ 𝜇̂ 𝛼̂ 𝜅̂ 𝜇̂ 𝛼̂ 𝜇̂ 𝛼̂ 𝜅̂ 𝜇̂ 𝛼̂ 𝜅̂

16722 1993.73 450.20 0.16 297.25 128.76 14.72 1955.47 433.50 2149.33 280.04 − 0.11 2152.08 483.32 − 0.17
(48.38) (34.34) (0.07) (763.60) (58.50) (12.50) (43.91) (33.52) (49.03) (22.95) (0.07) (51.38) (33.78) (0.09)
2012 768.35 223.53 0.43 1663.77 − 52.30 15.99 718.05 240.19 839.62 117.07 0.08 844.26 204.85 0.16
(34.29) (24.67) (0.08) (516.20) (36.45) (20.76) (36.57) (25.24) (29.98) (13.92) (0.10) (31.44) (21.03) (0.12)
2357 207.26 52.41 0.24 − 285.00 5.69 90.04 200.73 50.87 225.42 30.92 − 0.04 225.60 53.77 − 0.06
(9.80) (6.86) (0.11) (869.20) (9.86) (308.12) (9.01) (6.84) (9.49) (4.35) (0.12) (9.93) (6.45) (0.15)
4 281.69 74.13 0.02 113.75 40.43 5.18 280.81 73.61 307.58 48.63 − 0.20 308.92 85.91 − 0.32
(11.84) (8.10) (0.11) (51.00) (15.39) (3.10) (10.49) (8.00) (11.97) (6.16) (0.08) (12.80) (9.24) (0.12)
957 108.80 26.47 0.08 61.85 17.66 3.42 107.71 25.83 119.34 17.35 − 0.13 118.11 29.65 − 0.28
(5.18) (3.77) (0.14) (27.44) (13.39) (4.08) (4.56) (3.47) (5.56) (2.54) (0.14) (5.76) (3.86) (0.19)
Station Probability distributions
GEV P3 GUM GLO LN3
LM LM LM LM LM
𝜇̂ 𝛼̂ 𝜅̂ 𝜇̂ 𝛼̂ 𝜅̂ 𝜇̂ 𝛼̂ 𝜇̂ 𝛼̂ 𝜅̂ 𝜇̂ 𝛼̂ 𝜅̂

16722 1990.07 456.59 0.15 11.28 114.04 19.12 1959.44 403.29 2157.98 276.98 − 0.07 2154.46 490.66 − 0.15
(46.62) (36.53) (0.07) (994.14) (57.03) (19.50) (40.39) (36.53) (46.90) (22.79) (0.05) (48.38) (35.96) (0.09)
2012 764.68 220.31 0.40 1964.77 − 39.13 29.07 728.59 170.98 839.06 117.80 0.06 840.27 208.73 0.12
(33.65) (23.58) (0.11) (781.80) (42.25) (22.81) (26.28) (22.34) (28.85) (13.84) (0.08) (32.31) (22.61) (0.13)
2357 206.33 53.49 0.22 − 305.49 5.80 91.88 201.21 45.18 225.54 31.26 − 0.03 225.36 55.39 − 0.07
(10.49) (6.96) (0.13) (270.70) (15.19) (23.56) (8.01) (7.06) (9.51) (4.46) (0.09) (10.37) (6.62) (0.16)
4 279.83 73.16 − 0.01 148.25 50.90 3.43 280.25 74.02 308.19 48.68 − 0.18 306.66 85.97 − 0.37
(10.94) (9.12) (0.10) (89.07) (21.98) (8.62) (10.75) (9.07) (12.16) (6.18) (0.08) (13.10) (9.70) (0.14)
957 109.40 28.50 0.14 − 3.65 7.84 16.06 107.62 25.38 119.92 17.40 − 0.08 119.68 30.82 − 0.17
(5.50) (4.18) (0.13) (103.66) (7.83) (20.99) (4.55) (4.10) (5.26) (2.54) (0.10) (5.90) (3.93) (0.17)

Here, 𝜇̂ , 𝛼̂ and 𝜅̂ represent the estimated location, scale and shape parameters, respectively. The value in parenthesis is a standard error of
estimated parameter

the LM method are best-fit distributions for gauging site
Kukkolankoski Övre and Övre Abiskojokk, respectively.
In this study, a single distribution has not emerged as
the best distribution for all gauging sites. This was also
the case in [1, 5]. Overall, the LM estimation method per-
formed better for identifying the suitable distribution (also
see, [1]). The most suited distribution with MLE estimation
method is identified at gauging site which has the high-
est CV and skewness, see Tables 5 and 6. It seems that the
sites having extreme average of annual maxima of flood
and catchment area (either very large or very small) are in
favour of the LM method of estimation, see Tables 1, 5 and
6. If we look the landscape setting, the gauging sites which
are at an extreme position (close and far away) to the Gulf

Vol:.(1234567890)
SN Applied Sciences (2019) 1:1629 | https://doi.org/10.1007/s42452-019-1584-z Research Article

Table 5  Descriptive statistics Station n Mean Median S CV Skewness Kurtosis

(cubic metre per second)
16722 108 2192.22 2140.00 493.03 0.22 0.37 − 0.19
2012 50 827.28 837.53 211.16 0.26 − 0.54 0.90
2357 35 227.29 223.53 54.71 0.24 0.17 − 0.07
04 52 322.98 296.50 93.32 0.29 0.93 0.95
957 34 122.27 120.67 31.39 0.26 0.73 1.24

Here, S represents the standard deviation and CV indicates the coefficient of variation

of Bothnia are in favour of the LM estimation method.
The sample size of the time series does not seem to be
an important factor in favour of particular distribution or
estimation method in this study.
One major objective of flood frequency analysis is to
estimate the quantiles in the extreme upper tail of the
best-fitted distribution at each gauging site. The quantiles
estimate for the return periods 5, 10, 25, 50, 100, 200 and
500 years are calculated by using quantile function and
parameters value of the best-fitted distributions. Quan-
tile estimate yT with non-exceedance probability F for the
best-fitted distributions is given in Table 7. The estimate of
uncertainty ( 𝜎s ) in quantile estimates and 95% confidence
intervals of quantiles of flood for different return period
are also presented in Table 7. The SE indicates that longer
return periods have more uncertainty around the flood
quantile estimates.
5 Conclusion
In this study, the annual maximum steam flow series of
five gauging sites of Torne River are examined. Flood
frequency analysis is performed by using GEV, P3, GUM,
GLO and LN3 distributions. The MLE and LM parameter
estimation techniques are used to estimate the distribu-
tion’s parameters. The study investigates the selection of
best-fit probability distribution and estimation method for
at-site flood frequency analysis of Torne river. The best-fit
frequency distribution is identified at each gauging site
based on the highest total rank score in goodness-of-fit
tests and accuracy measures.
The results indicate that the GLO distribution using MLE
for gauging site Junosuando and the LN3 distribution with
a LM method for Pajala Pumphus and Abisko perform bet-
ter than other distributions of this study. The GEV and P3
distributions using the LM method are the most suitable
distribution at Kukkolankoski Övre and Övre Abiskojokk,
respectively. At most gauging sites, the best distributions
using LM estimation method are identified as the best-
suited distributions.
The results found in this research study for flood fre-
quency analysis of Torne River can be used in flood study,
water resource planning and designing of hydraulic
structures within the same basin and similar catchments.
The best-fitted distributions used in this study could be
considered as candidate distributions for regional flood
frequency analysis of Torne River basin or at-site flood fre-
quency analysis on other rivers in Sweden as well.

Vol.:(0123456789)
Research Article SN Applied Sciences (2019) 1:1629 | https://doi.org/10.1007/s42452-019-1584-z

Table 6  Rank score of Station Method Distribution AD RMSE MAE RMSEP MAEP Total rank
R2
distribution in both GOF tests
and accuracy measures Kukkolankoski Övre MLE GEV 7 5 7 7 7 5 38
P3 6 7 6 6 5 7 37
GLO 4 3 4 4 4 3 22
LN3 5 6 5 5 6 6 33
GUM 2 1 2 2 2 1 10
LM GEV 10 10 10 10 10 10 60
P3 9 9 9 9 9 9 54
GLO 3 4 3 3 3 4 20
LN3 8 8 8 8 8 8 48
GUM 1 2 1 1 1 2 08
Pajala Pumphus (2012) MLE GEV 3 4 7 4 4 4 26
P3 6 5 5 5 5 5 31
GLO 4 8 3 8 8 8 39
LN3 7 6 6 6 6 6 37
GUM 1 1 1 2 1 1 7
LM GEV 5 3 8 3 3 3 25
P3 9 9 9 9 9 9 54
GLO 8 7 4 7 7 7 40
LN3 10 10 10 10 10 10 60
GUM 2 2 2 1 2 2 11
Abisko (2357) MLE GEV 3 3 3 3 4 3 19
P3 8 5 6 4 6 5 34
GLO 4 6 4 9 3 6 32
LN3 7 4 5 5 5 4 30
GUM 2 1 2 2 2 1 10
LM GEV 6 7 8 6 8 7 42
P3 9 9 9 7 9 9 52
GLO 5 8 7 10 7 8 45
LN3 10 10 10 8 10 10 58
GUM 1 2 1 1 1 2 8
Junosuando (4) MLE GEV 5 1 1 2 2 1 12
P3 2 1 1 2 2 1 09
GLO 10 10 6 9 7 10 52
LN3 4 3 4 5 5 3 24
GUM 8 6 9 6 8 6 43
LM GEV 6 9 8 7 6 9 45
P3 1 4 5 1 1 4 16
GLO 9 5 3 10 9 5 41
LN3 3 7 7 4 4 7 32
GUM 7 8 10 8 10 8 51
Övre Abiskojokk (957) MLE GEV 6 5 7 7 7 5 37
P3 1 7 1 4 1 7 21
GLO 5 10 4 3 4 10 36
LN3 4 6 5 6 6 6 33
GUM 3 9 3 5 3 9 32
LM GEV 8 1 8 9 8 1 35
P3 9 2 10 10 10 2 43
GLO 7 4 6 1 5 4 27
LN3 10 3 9 8 9 3 42
GUM 2 8 2 2 2 8 24

The bold values are rank score for best-fitted distribution

Vol:.(1234567890)
SN Applied Sciences (2019) 1:1629 | https://doi.org/10.1007/s42452-019-1584-z Research Article

Table 7  Quantile estimates of flood with 95% confidence interval at five gauging sites of Torne River

Station Distribution Method Statistics Non-exceedance probability and return periods (years)
0.80 0.90 0.96 0.98 0.99 1.00 1.00
5.00 10.00 25.00 50.00 100.00 200.00 500.00

16772 GEV LM Lower 2475.06 2703.84 2929.24 3056.12 3157.22 3236.94 3324.16
Fit 2601.25 2857.74 3142.00 3327.52 3492.74 3640.52 3812.68
Upper 2728.22 3011.45 3358.67 3612.61 3867.51 4120.59 4461.08
𝜎s 64.33 78.28 109.83 142.87 181.92 225.49 288.13
2012 LN3 LM Lower 943.17 1016.74 1080.04 1114.79 1141.18 1164.31 1188.27
Fit 1007.08 1087.59 1168.75 1218.72 1262.09 1300.52 1345.53
Upper 1067.68 1155.55 1258.97 1332.30 1402.33 1470.83 1558.80
𝜎s 31.69 35.55 45.80 55.86 66.96 78.66 94.65
2357 LN3 LM Lower 250.78 272.75 293.06 304.31 313.38 321.13 330.25
Fit 273.37 299.61 328.49 347.65 365.24 381.63 401.88
Upper 296.48 327.91 368.54 400.46 433.38 466.35 511.51
𝜎s 11.66 14.15 19.52 24.73 30.66 37.15 46.49
04 GLO MLE Lower 350.20 392.09 446.76 486.07 529.94 577.23 641.49
Fit 385.52 442.47 525.23 596.78 678.28 771.63 916.74
Upper 424.56 499.39 629.33 754.07 920.40 1132.33 1491.74
𝜎s 18.70 27.11 46.15 68.45 99.73 142.66 223.51
957 P3 LM Lower 133.74 146.98 159.11 166.19 171.67 176.38 181.46
Fit 147.67 163.85 182.29 194.88 206.65 217.81 231.83
Upper 162.08 183.06 210.76 231.51 251.71 272.35 299.50
𝜎s 7.19 9.28 13.09 16.44 20.02 23.77 28.89

The lower limit is the fifth percentile and upper limit is the 95th percentile of the quantile distribution. The standard deviation of a quantile
distribution is 𝜎s which represents the standard error of quantile estimates. The flood quantile estimates for different return periods indicate
with bold value

Acknowledgements  Open access funding provided by Stockholm
University.
Data availability  The data that support the findings of this study are
openly available at SMHI [25].
Compliance with ethical standards  5. Drissia TK, Jothiprakash V, Anitha AB (2019) Flood frequency
Conflict of interest  The authors declare that they have no conflict of
interest.
Open Access  This article is distributed under the terms of the Crea-
tive Commons Attribution 4.0 International License (http://creati​ veco​
mmons​.org/licen​ses/by/4.0/), which permits unrestricted use, distri-
bution, and reproduction in any medium, provided you give appro-
priate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made.
References
1. Ahmad I, Fawad M, Mahmood I (2015) At-site flood frequency
analysis of annual maximum stream flows in Pakistan using
robust estimation methods. Pol J Environ Stud 24(6):2345–2353
2. Castellarin A, Kohnová S, Gaál L, Fleig A, Salinas JL, Toumazis A,
Kjeldsen TR, Macdonald N (2012) Review of applied-statistical
methods for flood-frequency analysis in Europe. Technical
report, (NERC) Centre for Ecology & Hydrology
3. Cicioni G, Giuliano G, Spaziani FM (1973) Best fitting of prob-
ability functions to a set of data for flood studies. In: Floods and
droughts
4. Cunnane C (1989) Statistical distributions for flood frequency
analysis. Operational Hydrology Report (WMO)
analysis using L moments: a comparison between at-site and
regional approach. Water Resour Manag 33(3):1013–1037
6. Elfvendahl S, Liljaniemi P, Salonen N (2006) The River Torne inter-
national watershed: common Finnish and Swedish typology,
reference conditions and a suggested harmonised monitoring
program: results from the TRIWA project. County Administrative
Board of Norrbotten [Länsstyrelsen i Norrbottens län]
7. Haddad K, Rahman A (2011) Selection of the best fit flood fre-
quency distribution and parameter estimation procedure: a case
study for Tasmania in Australia. Stoch Environ Res Risk Assess
25(3):415–428
8. Hamed K, Rao AR (1999) Flood frequency analysis. CRC Press, Boca
Raton
9. Hosking JRM (1986) The theory of probability weighted moments.
IBM Research Rep RC12210, IBM, Yorktown Heights, NY Google
Scholar
10. Hosking JRM (1990) L-moments: analysis and estimation of distri-
butions using linear combinations of order statistics. J R Stat Soc
Ser B (Methodol) 52(1):105–124

Vol.:(0123456789)
Research Article SN Applied Sciences (2019) 1:1629 | https://doi.org/10.1007/s42452-019-1584-z
11. Kite GW (2019) Frequency and risk analyses in hydrology.
Water Resour Publications, LLC. https​://books​.googl​e.se/books​
?id=b9OKx​AEACA​AJ
12. Laio F (2004) Cramer–von Mises and Anderson–Darling goodness
of fit tests for extreme value distributions with unknown param-
eters. Water Resour Res 40(9):W09308
13. Leandro J, Leitão JP, de Lima JLMP (2013) Quantifying the uncer-
tainty in the soil conservation service flood hydrographs: a case
study in the Azores Islands. J Flood Risk Manag 6(3):279–288
14. Leandro J, Gander A, Beg MNA, Bhola P, Konnerth I, Willems W, Car-
valho R, Disse M (2019) Forecasting upper and lower uncertainty
bands of river flood discharges with high predictive skill. J Hydrol
576:749–763
15. Merz B, Thieken AH (2005) Separating natural and epistemic
uncertainty in flood frequency analysis. J Hydrol 309(1–4):114–132
16. Meylan P, Favre AC, Musy A (2012) Predictive hydrology: a fre-
quency analysis approach. CRC Press, Boca Raton
17. Mkhandi SH, Kachroo RK, Gunasekara TAG (2000) Flood frequency
analysis of Southern Africa: II. Identification of regional distribu-
tions. Hydrol Sci J 45(3):449–464
18. Młyński D, Wałęga A, Stachura T, Kaczor G (2019) A new empirical
approach to calculating flood frequency in ungauged catchments:
a case study of the upper Vistula basin, Poland. Water 11(3):601
19. Nelder JA, Mead R (1965) A simplex method for function minimiza-
tion. Comput J 7(4):308–313
Vol:.(1234567890)
20. Önöz B, Bayazit M (1995) Best-fit distributions of largest available
flood samples. J Hydrol 167(1–4):195–208
21. Opere AO, Mkhandi S, Willems P (2006) At site flood frequency
analysis for the Nile Equatorial basins. Phys Chem Earth Parts
A/B/C 31(15–16):919–927
22. Rahman AS, Rahman A, Zaman MA, Haddad K, Ahsan A, Imteaz M
(2013) A study on selection of probability distributions for at-site
flood frequency analysis in Australia. Nat Hazards 69(3):1803–1813
23. Saf B (2009) Regional flood frequency analysis using L-moments
for the West Mediterranean region of Turkey. Water Resour Manag
23(3):531–551
24. Sevruk B, Geiger H (1981) Selection of distribution types for
extremes of precipitation (No. 551.577). Secretariat of the World
Meteorological Organization
25. The Swedish Meteorological and Hydrological Institute (2019)
Hydrologiska observationer. Data files retrieved from SMHI
hydrological observations. https​://vatte​nwebb​.smhi.se/stati​on/.
Accessed 20 Mar 2019
Publisher’s Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.